Studies on solid state reactions of atomic layer deposited thin ......Studies on solid state reactions of atomic layer deposited thin ﬁlms of lithium carbonate with hafnia and zirconia

This is an electronic reprint of the original article.This reprint may differ from the original in pagination and typographic detail.

Powered by TCPDF (www.tcpdf.org)

This material is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user.

Mäntymäki, Miia; Atosuo, Elisa; Heikkilä, Mikko; Vehkamäki, Marko; Mattinen, Miika;Mizohata, Kenichiro; Räisänen, Jyrki; Ritala, Mikko; Leskelä, MarkkuStudies on solid state reactions of atomic layer deposited thin films of lithium carbonate withhafnia and zirconia

Published in:JOURNAL OF VACUUM SCIENCE AND TECHNOLOGY A

DOI:10.1116/1.5081494

Published: 01/03/2019

Document VersionPublisher's PDF, also known as Version of record

Published under the following license:Unspecified

Please cite the original version:Mäntymäki, M., Atosuo, E., Heikkilä, M., Vehkamäki, M., Mattinen, M., Mizohata, K., Räisänen, J., Ritala, M., &Leskelä, M. (2019). Studies on solid state reactions of atomic layer deposited thin films of lithium carbonate withhafnia and zirconia. JOURNAL OF VACUUM SCIENCE AND TECHNOLOGY A, 37(2), 020929. [020929].https://doi.org/10.1116/1.5081494

https://doi.org/10.1116/1.5081494https://doi.org/10.1116/1.5081494

Studies on solid state reactions of atomic layer deposited thin films of lithiumcarbonate with hafnia and zirconiaMiia Mäntymäki, Elisa Atosuo, Mikko J. Heikkilä, Marko Vehkamäki, Miika Mattinen, Kenichiro Mizohata, JyrkiRäisänen, Mikko Ritala, and Markku Leskelä

Citation: Journal of Vacuum Science & Technology A 37, 020929 (2019); doi: 10.1116/1.5081494View online: https://doi.org/10.1116/1.5081494View Table of Contents: https://avs.scitation.org/toc/jva/37/2Published by the American Vacuum Society

ARTICLES YOU MAY BE INTERESTED INToward epitaxial ternary oxide multilayer device stacks by atomic layer depositionJournal of Vacuum Science & Technology A 37, 020602 (2019); https://doi.org/10.1116/1.5081997Functional model for analysis of ALD nucleation and quantification of area-selective depositionJournal of Vacuum Science & Technology A 37, 020911 (2019); https://doi.org/10.1116/1.5054285Comparative study on the use of novel heteroleptic cyclopentadienyl-based zirconium precursors with H2O andO3 for atomic layer deposition of ZrO2Journal of Vacuum Science & Technology A 37, 020912 (2019); https://doi.org/10.1116/1.5079539Crystalline tungsten sulfide thin films by atomic layer deposition and mild annealingJournal of Vacuum Science & Technology A 37, 020921 (2019); https://doi.org/10.1116/1.5074153Thermal atomic layer etching of HfO2 using HF for fluorination and TiCl4 for ligand-exchangeJournal of Vacuum Science & Technology A 36, 061504 (2018); https://doi.org/10.1116/1.5045130Atomic layer deposition of cobalt(II) oxide thin films from Co(BTSA)2(THF) and H2OJournal of Vacuum Science & Technology A 37, 010908 (2019); https://doi.org/10.1116/1.5066638

Studies on solid state reactions of atomic layer deposited thin filmsof lithium carbonate with hafnia and zirconia

Miia Mäntymäki,1,a) Elisa Atosuo,1 Mikko J. Heikkilä,1 Marko Vehkamäki,1 Miika Mattinen,1

Kenichiro Mizohata,2 Jyrki Räisänen,2 Mikko Ritala,1 and Markku Leskelä11Department of Chemistry, University of Helsinki, P.O. Box 55, FI-00014 Helsinki, Finland2Department of Physics, University of Helsinki, P.O. Box 43, FI-00014 Helsinki, Finland

(Received 15 November 2018; accepted 11 February 2019; published 28 February 2019)

In this paper, results on the solid state reactions of atomic layer deposited Li2CO3 with HfO2 andZrO2 are reported. An Li2CO3 film was deposited on top of hafnia and zirconia, and the stackswere annealed at various temperatures in air to remove the carbonate and facilitate lithium diffu-sion into the oxides. It was found that Li+ ions are mobile in hafnia and zirconia at high tempera-tures, diffusing to the film–substrate interface and forming silicates with the Si substrate duringheating. Based on grazing incidence x-ray diffraction experiments, no changes in the oxide phasestake place during this process. Field emission scanning electron microscopy images reveal thatsome surface defects are formed on the transition metal oxide surfaces during lithium diffusion.The authors also show that lithium can diffuse through hafnia and react with a potential lithium-ion battery electrode material TiO2 residing below the HfO2 layer, forming Li2TiO3. Published bythe AVS. https://doi.org/10.1116/1.5081494

I. INTRODUCTION

Hafnium and zirconium oxides are well-known high-κmaterials (≈30 for HfO2, ≈25 for ZrO2),1 useful forMOSFETs and memory devices.2 Due to their high electricalresistivity, these oxides could prove useful also in all-solid-state lithium-ion batteries as electrolyte materials. However,electrolyte materials are also required to have a high enoughlithium-ion conductivity to be useful. The commonly usedsolid electrolyte LiPON shows Li-ion conductivities between10−6 and 10−8 S/cm.3 With the emergence of ever smaller,3D-structured batteries, geometric reasons could permit the uti-lization of even somewhat lower ionic conductivities in batterymaterials. Even though the utilization of hafnia and zirconia aselectrolytes in lithium-ion batteries has not been studied assuch, some experiments in this area have already been done.4,5

For example, it has been reported that ultrathin HfO2 layerscan be used to protect SnO2 nanoparticle anodes, resulting inan improved electrochemical operation of the anode.4 Evenafter 200 cycles of HfO2 (growth rate ≈1 Å/cycle), the batteryperformance is improved, without a loss in diffusion kinetics.These results indicate that at least through thin layers ofHfO2, lithium diffusion is fast enough for battery applica-tion. In addition, simple lithium-ion diffusion into ZrO2has also been studied.5 It was found that monoclinic ZrO2is a poor lithium-ion conductor and that lithium diffusivitydecreases as the amount of Li+ in the oxide increases.

In this paper, we have studied the interaction of Li2CO3thin films with HfO2 and ZrO2 thin films. Previously, wehave shown that it is possible to form lithium transition

metal oxide thin films from lithium carbonate and metaloxides.6 In the current experiments, our goal was to studylithium diffusion into hafnia and zirconia during heating.In addition, we hoped to find out whether lithium hafnatesor zirconates could be formed in this manner. Hafniumoxide does not have many well-known lithium containingphases, although at least the structures of two crystallinelithium hafnium oxides are known in the literature, namely,Li2HfO3 and Li8HfO6.

7,8 Lithium hafnate Li2HfO3 can beused, for example, as a scintillator material.9 Li6Hf2O7 hasalso been reported, but very little information on this mate-rial is given. However, based on solid-state Li NMRstudies, this material was concluded to be a poor Li-ionconductor.10 Lithium zirconates, on the other hand, aremuch better known in materials science.11–15 Li2ZrO3 hasbeen reported to be a possible tritium breeding material forfuture fusion reactors.14 In addition, it can be used toprotect lithium-ion battery cathode materials,11,12 and itsapplication as an anode material in lithium-ion batterieshas also been studied.13 Doped Li8ZrO6, on the other hand,has been suggested as a possible cathode material forLi-ion batteries.15,16

II. EXPERIMENT

A. Film deposition

Li2CO3 thin films were deposited in an ASMMicrochemistry F-120 hot-wall flow-type atomic layer deposi-tion (ALD) reactor onto ZrO2 and HfO2 thin films. These filmshad been deposited by ALD onto single-crystalline siliconwafers using a heteroleptic zirconium precursor and water17

and TEMAH [tetrakis(ethylmethylamino)hafnium] and water.1

Lithd (Volatec oy) and ozone, generated by a Wedeco GmbHModular 4 HC ozone generator, were used as precursorsfor the Li2CO3 depositions as previously described in the

Note: This paper is part of the 2019 special collection on Atomic LayerDeposition (ALD).a)Present address: Department of Chemistry and Materials Science, Schoolof Chemical Engineering, Aalto University, P.O. Box 16100, FI-00076Aalto, Finland; electronic mail: [email protected]

020929-1 J. Vac. Sci. Technol. A 37(2), Mar/Apr 2019 0734-2101/2019/37(2)/020929/8/$30.00 Published by the AVS. 020929-1

literature.18 The ozone flow rate was 30 l/h, and the concen-tration was 100 g/Nm3. Lithd was evaporated inside thereactor at 192 °C at a pressure of approximately 5 mbar. Thepulse time for Lithd was 1.5 s with a 2 s purge, and ozonewas pulsed for 2.5 s with a 3.5 s purge time. The pulsing ofthe lithium precursor was done by inert gas valving, withgaseous N2 as the pulse and purge gas. The N2, obtainedfrom liquid N2, had as an impurity less than 3 ppm of H2Oand O2 each. All the Li2CO3 deposition experiments weredone at 225 °C.

B. Film characterization

A muffle furnace was used for the annealing of the films.All annealing experiments were done in air with a maximumheating rate of approximately 9 °C/min.

The crystallinity of the films was studied by grazingincidence x-ray diffraction (GIXRD) measurements usinga PANalytical X’Pert Pro MPD x-ray diffractometer. Insitu high-temperature XRD (HTXRD) measurements werealso conducted with an Anton-Paar HTK1200N oven. Themorphology of the films before and after annealing wasstudied by field emission scanning electron microscopy

with a Hitachi S4800 FESEM instrument. For the FESEMimaging, the samples were coated with approximately 3nm of Au/Pd by sputtering. No coating was used in thecross-sectional FESEM analyses. Transmission electronmicroscope (TEM) specimens were prepared using stan-dard focused ion beam lift-out procedures. Bright-fieldTEM images were taken with an FEI Tecnai F20 micro-scope operated at 200 kV.

Surface roughness was quantified with atomic forcemicroscopy (AFM) using a Veeco Multimode V instrument.Tapping mode images were captured in air using siliconprobes with a nominal tip radius of 10 nm and a nominalspring constant of 5 N/m (Tap150 from Bruker). Imageswere flattened to remove artifacts caused by sample tilt andscanner bow. Film roughness was calculated as a root-mean-square value (Rq).

The composition of the films was studied with time-of-flight elastic recoil detection analysis (ToF-ERDA). TheToF-ERDA measurements were performed with 50MeV127I and 40MeV 79Br beams from the 5MV EGP-10-IItandem accelerator at the University of Helsinki.19 The detec-tion angle was 40° and the sample was tilted 15° relative tothe beam direction.

FIG. 1. High-temperature x-ray diffractograms of a 3000 cycle Li2CO3 filmdeposited onto 50 nm HfO2. The annealing was done in air. Patterns:Li2CO3 = PDF 22–1141, JCPDS-ICDD, HfO2 = PDF 34–0104,JCPDS-ICDD, Li2SiO3 = PDF 29–0828, JCPDS-ICDD.

FIG. 2. X-ray diffractograms of a 3000 cycle Li2CO3 film deposited onto50 nm HfO2 measured as-deposited and after annealing in air at 650 °C for2 h. Patterns: Li2CO3 = PDF 22–1141, JCPDS-ICDD, HfO2 = PDF 34–0104,JCPDS-ICDD, Li2SiO3 = PDF 29–0828, JCPDS-ICDD.

FIG. 3. FESEM images of 3000 cycles of Li2CO3 deposited by ALD onto 50 nm of HfO2, before (a) and after (b) annealing in air at 650 °C.

020929-2 Mäntymäki et al.: Studies on solid state reactions of atomic layer deposited thin films of Li2CO3 020929-2

J. Vac. Sci. Technol. A, Vol. 37, No. 2, Mar/Apr 2019

III. RESULTS AND DISCUSSION

A. HfO2 films as substrates

To study whether lithium containing hafnium oxidescould be made by similar solid state reactions as we havereported for Li2TiO3, LiTaO3, and LiNbO3,

6 a 50 nmhafnium oxide film was covered with 3000 cycles of atomiclayer deposited Li2CO3 and the resulting stack was subjectedto a high-temperature XRD measurement (Fig. 1). In thetemperature range studied, no crystalline lithium hafnatephases formed. At 565 °C, peaks belonging to Li2CO3 havedisappeared. At the same time, a peak appears at 2θ = 22.2°.The peak is difficult to index, but it could belong to alithium silicate phase Li4SiO4.

20 Hafnium oxide remainsmonoclinic during heating. At 655 °C, orthorhombic lithiumsilicate Li2SiO3 has formed. At the same time, the peak at22.2° has decreased, possibly indicating a shift from a meta-stable silicate to a more stable one.

For comparison with previous results on other materials,6

the Li2CO3/HfO2-stack was annealed in air at 650 °C for 2 hand studied with grazing incidence x-ray diffraction at roomtemperature (Fig. 2). As with the HTXRD measurement, noLi2CO3 or lithium hafnates could be seen after the annealing,with HfO2 remaining monoclinic. For the HfO2 layer, theunit cell parameters change 0.2%–0.4% upon annealing,with a 0.05° change in the β-angle. The literature values forthese parameters for the monoclinic HfO2 lie between theresults for the as-deposited and the annealed films, meaningthat the actual change in the HfO2 structure is smaller than0.4% despite lithium diffusing through the film. A small

increase in the crystallite size can be deduced from theresults. Based on the wide peaks visible especially in theas-deposited HfO2 films, the hafnia crystallites reside in anamorphous oxide matrix. After annealing, Li2SiO3 is presentin the stack, with no indication of other silicate phases.

The films were imagined with FESEM to study theirmorphology (Fig. 3). The as-deposited film showed largeplatelets on the surface, as is common for Li2CO3 films.

18

After the annealing, the film showed much smaller, column-like crystallites, and a smoother surface in general, somewhatsimilarly to titanium oxide films reacted with Li2CO3 inour previous study.6 Atomic layer deposited HfO2 has beenreported to show columnlike crystallites after annealing.21

The change in roughness was quantified with AFM measure-ments (Fig. 4). For the as-deposited film, the rms roughnesswas as high as 47 nm, with large platelets comprising thesurface. After annealing, the roughness of the film stackdecreased to 2.9 nm, which is a typical value for polycrystal-line ALD HfO2 films.

1

In the FESEM images, the annealed film showed somecracking on the surface. This differs from the results obtainedfor other films studied previously in this same manner.6 Oneexplanation for the surface defects could be the reaction with

FIG. 4. AFM images of Li2CO3/HfO2-stacks before (a) and after (b) annealingin air at 650 °C.

TABLE I. ToF-ERDA results (at. %) of a 3000 cycle Li2CO3 film depositedonto 50 nm HfO2 and measured as-deposited and after annealing at 650 °Cfor 2 h in air.

Li2CO3/HfO2 as-depositedLi2CO3/HfO2 afteranneal at 650 °C

Li 18.9 ± 0.4 22.0 ± 0.4Hf 11.5 ± 0.1 10.9 ± 0.2O 56.1 ± 0.5 63.8 ± 0.5C 9.7 ± 0.2 0.94 ± 0.07H 2.5 ± 0.5 2.2 ± 0.5F 1.25 ± 0.07 0.07 ± 0.03Na 0.06 ± 0.03

the silicon substrate, which could cause strain in the HfO2film. Some damage could also be caused by carbon dioxideleaving the film structure, even though this type of damagewas not evident in the films forming intermediate lithiumtransition metal oxide phases. In addition, adhesion problemscannot be completely ruled out, as the samples needed to becut for the FESEM imaging.

ToF-ERDA measurements revealed that the as-depositedfilm was close to stoichiometric HfO2 and Li2CO3 (Table I).After annealing, the film composition was approximately“Li2HfO6,” indicating a large excess of oxygen. This excesscould be explained by the silicate formation during anneal-ing. Possibly some additional SiO2 is forming during anneal-ing as a result of the high oxygen diffusivity in HfO2,

22 andthe forming silicon dioxide is further forming silicates.In addition to the lithium silicate seen in the XRD measure-ments, possibly also some hafnium silicates are forming: ahafnium silicate interfacial layer has been reported to formduring annealing of ultrathin HfO2 layers on silicon.

23 TheToF-ERDA depth profiles corroborate the XRD results inthat lithium is indeed concentrated on the film–Si substrateinterface after the annealing (Fig. 5). Only a small amount oflithium is incorporated into the HfO2 film. Moreover, unlikemost lithium containing systems, lithium is not concentratedon the outermost film surface.24,25

To study whether it would be possible to mix the lithiumions with hafnia more completely and without silicate forma-tion, a series of annealing experiments was made. 3000 cyclesof Li2CO3 was deposited onto 50 nm of HfO2, and the stackwas annealed in air at 300, 400, 500, and 600 °C for 2 h.In addition, samples were annealed at 300 °C for 4 h to seewhether a longer annealing time could make a difference inthe lithium diffusion. After annealing, the films were studiedwith both GIXRD and ToF-ERDA and the results are col-lected in Figs. 6 and 7. From Fig. 6, it is apparent that crys-talline lithium carbonate has reacted completely only afterannealing at 600 °C. Comparing this information withFig. 7 reveals that the lower temperatures used were notenough to promote full mixing of lithium and hafnium oxide.Some migration has occurred already at 400 and 500 °C, but

FIG. 7. ToF-ERDA depth profiles of an Li2CO3/HfO2 film stack after anneal-ing in air at 300 °C for 4 h (a), 400 °C for 2 h (b), 500 °C for 2 h (c), and600 °C for 2 h (d).

FIG. 8. FESEM and TEM images of a film stack composed of Li2CO3, HfO2, and TiO2 before [FESEM: (a), TEM: (c)] and after annealing at 650 °C for 4 h inair [FESEM: (b), TEM: (d)]. After the annealing, only two layers remain.



in these films, lithium is still strongly enriched on the stacksurface as the carbonate. In addition, the amount of carbonhas not decreased in these samples. Only after annealing at600 °C has the carbonate been decomposed, and lithiumhas diffused through the hafnia film to the silicon substrateinterface. During the annealing, the carbon impurities wereremoved from the film. However, this change was accompa-nied by an increase in the amount of hydrogen, from 2.2 at. %at 500 °C to 6.3 at. % at 600 °C. This could be explained byLiOH formation caused by lithium enrichment on the filmsurface and reaction with ambient air, similarly to the LiTaO3sample studied previously.6 Interestingly, despite the largeamount of lithium in the silicon substrate surface after theannealing at 600 °C, no crystalline lithium silicates can befound in Fig. 6 at this temperature.

With these experiments, we were able to demonstrate thatat elevated temperatures lithium ions can move through thewell-known insulating material HfO2 and react with the under-lying silicon dioxide and silicon. To show that the reaction isnot only limited to the formation of lithium silicates, a sampleshown in Fig. 8 was prepared. 3000 cycles of Li2CO3 wasdeposited onto 50 nm of HfO2, which had been depositedonto approximately 60 nm of TiO2. The TiO2 film wasdeposited using titanium(IV) iso-propoxide and water. Thethree-layer stack was annealed in air at 650 °C for 30 min,and 2 and 4 h. The x-ray diffractograms (Fig. 9) show thatalready after 30 min of annealing, the lithium ions had dif-fused through the hafnia into the TiO2 layer, forming crys-talline Li2TiO3. Some anatase is also present in this layer,based on both Fig. 9 and the metal ratio Li:Ti = 1.4:1, mea-sured with ToF-ERDA. Again, the HfO2 stays monoclinicbefore and after the annealing. Notably, no lithium silicatephases are detected with XRD. ToF-ERDA depth profilescorroborate the diffraction results by showing that after theannealing lithium is confined to the same layer togetherwith titanium (Fig. 10). In addition, both the depth profileand the FESEM image of the annealed film reveal that the

hafnia and titania layers have not mixed, but remain sepa-rate despite the lithium diffusion (Figs. 8 and 10).

The formation of the Li2TiO3 layer was also studied withTEM [Figs. 8(c) and 8(d)]. Prior to the annealing of theLi2CO3/HfO2/TiO2 film stack, a native silicon oxide layer ispresent in the TiO2/Si interface [Fig. 8(c)]. The anatase layeris comprised of a single layer of columnar grains. Severalstructural changes are observed in the annealed specimen[Fig. 8(d)]. Upon annealing, the apparent HfO2 layer thick-ness increases from 50 to ca. 58 nm. This can be at least inpart explained by the increased roughness of both the layersurface and the HfO2/Li2TiO3 interface. It is noteworthy thatthe HfO2 layer remains continuous, with no cracking orstrong mixing with the Li2TiO3 layer taking place. In theTEM image taken after the annealing step, the Li2TiO3 layerthickness is in the range of 80–89 nm, as measured fromindividual spots along the cross section, corresponding to ca.50% volume expansion compared to the original TiO2 layer.This is consistent with the expected molar volume expansionupon conversion of anatase TiO2 to Li2TiO3. The Li2TiO3

FIG. 10. ToF-ERDA depth profiles of an Li2CO3/HfO2/TiO2 stack before (a)and after annealing at 650 °C for 2 h in air (b).

FIG. 11. X-ray diffractograms of a 3000 cycle Li2CO3 film deposited onto54 nm ZrO2 and measured as-deposited and after annealing in air at 500 and650 °C. Patterns: Li2CO3 = PDF 22–1141, JCPDS-ICDD, ZrO2 = InorganicCrystal Structure Database (ICSD), collection code 66781, and PDF37–1484 JCPDS-ICDD, Li2SiO3 = PDF 29–0828, JCPDS-ICDD.

FIG. 9. X-ray diffractograms of an Li2CO3/HfO2/TiO2 stack before and afterannealing at 650 °C in air. Patterns: Li2CO3 = PDF 22–1141, JCPDS-ICDD,HfO2 = PDF 34–0104, JCPDS-ICDD, TiO2 = PDF 21–1272, JCPDS-ICDD,Li2TiO3 = PDF 33–0831, JCPDS-ICDD.


JVST A - Vacuum, Surfaces, and Films

layer has two distinct grain heights, a thicker layer with ca.60 nm tall grains in contact with the HfO2 layer and a lowerca. 25 nm layer in contact with the Si interface. This can bedue to the disappearance of the native SiO2 layer, part ofwhich may have been incorporated into the lower part of theLi2TiO3 layer. A 10–15 nm reaction layer, possibly lithiumdoped Si or amorphous Li silicate, can be seen in theLi2TiO3, with some isolated grains of material which havegrown through the mostly smooth Li2TiO3/Si interface layer.

B. ZrO2 films as substrates

Similarly to the study on HfO2, 3000 cycles of Li2CO3was deposited onto a 54 nm ZrO2 film at 225 °C, and theresulting film stack was annealed at various temperatures inair for 2 h. The x-ray diffractogram (Fig. 11) shows that at500 °C some Li2CO3 is still present in the film stack.However, at 650 °C, lithium has moved through the ZrO2film, forming silicates with silicon from the single-crystallinesilicon substrate. At the same time, ZrO2 has changed fromthe tetragonal phase of the as-deposited film to a mixture oftetragonal and monoclinic phases. No lithium zirconatephases formed in these conditions.

FESEM imaging was used to study the morphology of theLi2CO3/ZrO2-stack before and after the annealing at 650 °C(Fig. 12). The as-deposited film shows the same rough, flakysurface as in the case of Li2CO3 on the hafnium oxide films.After the annealing, the surface roughness is decreaseddramatically, again similarly as on hafnia. However, thesurface shows more defects and is also broken in someplaces [Fig. 12(c)]. The reason for this difference betweenhafnia and zirconia is difficult to explain. One possibility

could be that the strain caused by lithium insertion is largerin the zirconia film. Another possibility is that during theannealing impurities such as hydroxyls are leaving the zir-conia film, causing crater formation and cracks. HfO2 filmsdeposited using TEMAH and water have been reported tocontain only very minor amounts of carbon and hydrogenimpurities,1 making crater formation less likely in thismaterial. To study this problem further, we also imaginedan annealed ZrO2 film without Li2CO3 deposition. Thisfilm did not show similar cracking (not shown here), indi-cating that impurities in ZrO2 most likely do not completelyexplain the cracking of the film. It appears that lithium plays apart in the changes in the morphology of the annealed stack.

FIG. 12. FESEM images of 3000 cycles of Li2CO3 deposited by ALD onto 54 nm of ZrO2, before (a) and after [(b) and (c)] annealing in air at 650 °C.

TABLE II. ToF-ERDA results (at. %) of a 3000 cycle Li2CO3 film depositedonto 54 nm ZrO2 and measured as-deposited and after annealing at 500 and650 °C for 2 h in air.

Li2CO3/ZrO2as-deposited

Li2CO3/ZrO2 afteranneal at 500 °C

Li2CO3/ZrO2 afteranneal at 650 °C

Li 19.7 ± 1.0 20.2 ± 1.1 25.7 ± 1.3Zr 9.4 ± 0.3 9.3 ± 0.3 9.3 ± 0.3O 52.0 ± 1.0 54.3 ± 1.1 61.2 ± 1.2C 11.9 ± 0.5 11.4 ± 0.6 0.93 ± 0.15H 6.2 ± 0.9 4.5 ± 0.7 2.5 ± 0.6F 0.50 ± 0.09 0.21 ± 0.06 0.09 ± 0.04Cl 0.13 ± 0.04 0.06 ± 0.03 0.07 ± 0.03N 0.13 ± 0.05 0.11 ± 0.05 0.09 ± 0.05Elementalratios

Assumingstoichiometric ZrO2:Li:C:O = 1.7:1:2.8

Li:Zr:O = 2.2:1:5.8 Li:Zr:O = 2.8:1:6.6



ToF-ERDA results (Table II) show that Li2CO3 formsclose to stoichiometric onto ZrO2. Most likely, the slightdeviation from stoichiometry is caused by the calculation nottaking into account carbon impurities in the ZrO2 film,which were measured to be of the order of 2.4 at. % beforeannealing. It is apparent from Table II that annealing the filmstack at 500 °C for 2 h is not enough to remove the carbonand hydrogen impurities and to cause lithium migration.However, after annealing at 650 °C, the lithium migrationhas taken place, with carbon and hydrogen impuritiesdecreasing at the same time. The reaction is evident also inthe ToF-ERDA depth profiles (Fig. 13). Lithium has movedinto the silicon substrate after annealing at 650 °C, while atthe lower annealing temperature, the profile for the mostpart resembles that of the as-deposited sample. The depthprofile of the as-deposited sample also revealed that most ofthe hydrogen impurities reside in the zirconia layer. Despitebeing embedded in this bottom layer, these impurities canbe removed with annealing.

Based on these results, it is clear that hafnium and zirco-nium oxides behave quite similarly when annealed with alithium carbonate film on top. This is somewhat surprisingconsidering the large amount of literature on lithium zirco-nates: one could assume that producing lithium zirconates

with solid state reactions would be more straightforward thanobtaining lithium hafnates. To still study whether we couldsynthesize lithium zirconates with this solid state reactionprocess, a thinner ZrO2 film of 30 nm was used, togetherwith 3000 cycles of Li2CO3. The idea behind this experimentwas to drive lithium into the thinner zirconia layer at a lowertemperature than above so that lithium would not react withthe substrate but instead with zirconia, producing a crystal-line lithium zirconate. Figure 14 shows a ToF-ERDA depthprofile and an x-ray diffractogram of a film stack depositedat 225 °C and annealed at 500 °C. It is evident that with thethinner zirconia film, lithium mixing occurs throughout thefilm already after annealing at 500 °C. This has to do withthe smaller thickness of the zirconia film promoting fasterlithium diffusion. At 500 °C, our previous sample was tetrag-onal, whereas the thinner sample showed peaks belonging toboth tetragonal and monoclinic zirconia, which might alsomean that the ionic conductivity is different in this sample.In any case, it is evident that only crystalline ZrO2 is presentin the diffractogram of the annealed sample. The atomic per-centages in the annealed film, as determined by ToF-ERDA,were Li 25,3, Zr 7.2, O 53.7, C 5.8, H 7.3, F 0.6, and N 0.1.The excess oxygen is most likely in the form of lithiumcarbonate, since not all carbon has been removed at this

FIG. 13. ToF-ERDA depth profiles of an Li2CO3/ZrO2 film stack as-deposited (a) and after annealing at 500 °C (b) and 650 °C (c) for 2 h in air.

FIG. 14. ToF-ERDA depth profile (a) and x-ray diffractogram (b) of an Li2CO3/ZrO2 film stack after annealing at 500 °C for 2 h in air. The ZrO2 film thicknesswas 30 nm, and Li2CO3 was applied for 3000 cycles. Patterns: ZrO2 = Inorganic Crystal Structure Database (ICSD), collection code 66781, and PDF 37–1484JCPDS-ICDD.


JVST A - Vacuum, Surfaces, and Films

temperature. Based on these results, the closest lithium zirco-nate phase possible would be either Li2ZrO3 or Li6Zr2O7.However, the phase cannot be unambiguously recognized inthe x-ray diffractogram in Fig. 14(b). Still, the small, broadpeak at 22.1°–22.9° (2θ) could originate from a zirconatephase.26–28 Further identification of the phase was impossibleat this time, which is why we believe that making lithiumzirconate films with this process is not feasible.

It has been reported that forming Li2ZrO3 from eitherLiOH or Li2CO3 and ZrO2 using solid state reactionsinvolves much higher temperatures than, for example, in thecase of Li2TiO3.

14,29 Therefore, we are faced with a similarproblem with Li2ZrO3 as we were previously with producingLi4Ti5O12:

6 the temperature range available for our system islimited due to lithium reactivity with silicon. The chemicalpotential of lithium silicate formation appears to drive thelithium migration through the zirconium oxide layer to theinterface with the substrate before any reaction with the zir-conia can occur.

IV. SUMMARY AND CONCLUSIONS

In this paper, we have studied the reactions betweenatomic layer deposited Li2CO3 and HfO2 or ZrO2. Afterannealing the films at 650 °C, lithium silicates have formedwith no indication of lithium hafnate or zirconate phases.Thus, it appears that these materials behave differently fromTiO2, Ta2O5, and Nb2O5, which we have studied previouslyand found to form lithium containing ternary phases.6 Afterannealing, the films showed a relatively smooth surface withsome defects. We postulate that these defects are formedwhen impurities are removed from the metal oxide layer andwhen lithium is inserted into the silicon substrate. Next,more research should be done on both the amorphous phasesand the different crystalline phases of HfO2 and ZrO2, as thephase could have a large effect on the lithium diffusivity.5

In addition, further attempts to deposit Li2ZrO3 should bemade using a lithium diffusion barrier under the ZrO2 layer,as Li2ZrO3 could also prove to be a useful lithium-ionbattery material.

ACKNOWLEDGMENTS

The authors would like to thank the electron microscopyunit of the Institute of Biotechnology at the University of

Helsinki for providing TEM instrument access. Sanni Seppäläis thanked for the ZrO2 depositions. Financial support fromASM Microchemistry Oy is gratefully acknowledged. Thiswork was also supported by the Finnish Centre of Excellencein Atomic Layer Deposition.

1K. Kukli, M. Ritala, T. Sajavaara, J. Keinonen, and M. Leskelä, Chem.Vap. Deposition 8, 199 (2002).

2V. A. Gritsenko, T. V. Perevalov, and D. R. Islamov, Phys. Rep. 613, 1(2016).

3H. Xia, H. L. Wang, W. Xiao, M. O. Lai, and L. Lu, Int. J. Surf. Sci. Eng.3, 23 (2009).

4N. Yesibolati, M. Shahid, W. Chen, M. N. Hedhili, M. C. Reuter,F. M. Ross, and H. N. Alshareef, Small 10, 2849 (2014).

5A. K. Jonsson, G. A. Niklasson, M. Ritala, M. Leskelä, and K. Kukli,J. Electrochem. Soc. 151, F54 (2004).

6E. Atosuo, M. Mäntymäki, K. Mizohata, M. J. Heikkilä, J. Räisänen,M. Ritala, and M. Leskelä, Chem. Mater. 29, 998 (2017).

7PDF 23–1183, JCPDS-ICDD, International Center for Diffraction Data.8PDF 26–0847, JCPDS-ICDD, International Center for Diffraction Data.9Ya. V. Baklanova, A. V. Ishchenko, T. A. Denisova, L. G. Maksimova,B. V. Shulgin, V. A. Pustovarov, and L. V. Viktorov, Opt. Mater. 34, 1037(2012).

10R. Czekalla and W. Jeitschko, Z. Anorg. Allg. Chem. 619, 2038(1993).

11Z. Miao, H. Ni, H. Zhang, C. Wang, J. Fang, and G. Yang, J. PowerSources 264, 147 (2014).

12Y. Xu, Y. Liu, Z. Lu, H. Wang, D. Sun, and G. Yang, Appl. Surf. Sci.361, 150 (2016).

13Y. Dong, Y. Zhao, H. Duan, and J. Huang, Electrochim. Acta 161, 219(2015).

14D. Cruz, H. Pfeiffer, and S. Bulbulian, Solid State Sci. 8, 470 (2006).15S. Huang, B. E. Wilson, B. Wang, Y. Fang, K. Buffington, A. Stein, andD. G. Truhlar, J. Am. Chem. Soc. 137, 10992 (2015).

16S. Huang, B. E. Wilson, W. H. Smyrl, D. G. Truhlar, and A. Stein, Chem.Mater. 28, 746 (2016).

17S. Seppälä, doctoral thesis (University of Helsinki, Finland) (to be published).18M. Putkonen, T. Aaltonen, M. Alnes, T. Sajavaara, O. Nilsen, andH. Fjellvåg, J. Mater. Chem. 19, 8767 (2009).

19J. Jokinen, J. Keinonen, P. Tikkanen, A. Kuronen, T. Ahlgren, andK. Nordlund, Nucl. Instr. Methods B 119, 533 (1996).

20PDF 34–1416, JCPDS-ICDD, International Center for Diffraction Data.21M.-Y. Ho et al., J. Appl. Phys. 93, 1477 (2003).22S. Ferrari and G. Scarel, J. Appl. Phys. 96, 144 (2004).23G. He, M. Liu, L. Q. Zhu, M. Chang, Q. Fang, and L. D. Zhang, Surf.Sci. 576, 67 (2005).

24E. Østreng, H. H. Sønsteby, T. Sajavaara, O. Nilsen, and H. Fjellvåg,J. Mater. Chem. C 1, 4283 (2013).

25V. Miikkulainen, O. Nilsen, M. Laitinen, T. Sajavaara, and H. Fjellvåg,RSC Adv. 3, 7537 (2013).

26PDF 33–0843, JCPDS-ICDD, International Center for Diffraction Data.27Inorganic Crystal Structure Database (ICSD), collection code 41321.28Inorganic Crystal Structure Database (ICSD), collection code 73835.29J. Ida, R. Xiong, and Y. S. Lin, Sep. Purif. Technol. 36, 41 (2004).



Studies on solid state reactions of atomic layer deposited thin films of lithium carbonate with hafnia and zirconiaI. INTRODUCTIONII. EXPERIMENTA. Film depositionB. Film characterization

III. RESULTS AND DISCUSSIONA. HfO2 films as substratesB. ZrO2 films as substrates

IV. SUMMARY AND CONCLUSIONSReferences

Documents

Studies on solid state reactions of atomic layer deposited thin ......Studies on solid state reactions of atomic layer deposited thin ﬁlms of lithium carbonate with hafnia and zirconia